BEFORE YOU BUY ANYTHING — INCLUDING OURS
Four things that will do more
for your sleep than our product.
We sell a sleep supplement, so read this with that in mind. The evidence for these four behaviours is stronger than the evidence for any ingredient in any jar, including ours. If you haven’t done them, a supplement is the wrong place to start — and we’d rather tell you that now than take your money and have it not work.
WHY TIMING IS THE LEVER
← swipe to see the whole chart →
Move your last coffee earlier than feels necessary
Caffeine's half-life averages about five hours, so a 200 mg coffee at 2 PM still leaves roughly 50 mg circulating at midnight — about half a cup, at lights-out. In a controlled trial, 400 mg taken a full six hours before bed cut objectively measured sleep by more than an hour while participants reported sleeping normally. That's the trap: falling asleep is the part caffeine disturbs least. A pooled analysis puts the cutoff for a standard cup at about 8.8 hours before bed.
TRY THIS
For two weeks, hold bedtime steady and cut caffeine at bedtime minus ten hours. Change nothing else.
Finish drinks three to four hours before bed
Alcohol shortens the time it takes to fall asleep at every dose tested — and then collects. The second half of the night gets fragmented, REM is delayed at every dose, and the classic 2-to-4 AM wake-up is roughly when your liver finishes clearing it. Wearable data from over 4,000 adults found even a single drink measurably reduced physiological recovery during the early hours of sleep, and neither being fit nor being young protected people.
TRY THIS
Same drinks, three hours earlier. The difference shows up in the back half of the night, which is where you feel it.
Get light in your eyes early, and take it out late
Specialised cells in your retina report light levels straight to your brain's master clock, most sensitively at the blue wavelengths screens and LED bulbs emit. Ordinary indoor room light before bed — not screens, just lamps — has been shown to suppress melatonin and shorten the body's internal night by around 90 minutes. The fix is a hardware-store trip, not a renovation: bright light within thirty minutes of waking, and the last two hours dim and warm.
TRY THIS
Ten minutes outside in the morning. Overhead lights off after dinner.
Make the room cold enough to need a blanket
Falling asleep is partly a cooling event — your core temperature has to drop before sleep begins, and a warm room fights that. Most sleep researchers land around 60–67°F. The counterintuitive part: warm feet help, because your body sheds core heat through its extremities. A warm bath one to two hours before bed works for the same reason — you cool down faster afterwards.
TRY THIS
Drop the thermostat two degrees. Socks.
RUN YOUR OWN NUMBERS
Your caffeine cutoff, mapped across the day.
Enter the dose, your last caffeine, bedtime, and the metabolism that feels closest. The dedicated calculator shows the remaining caffeine and the planning cutoff on one clear timeline.
Open the caffeine calculator →SO WHERE DOES A SUPPLEMENT FIT?
After the four, not instead of them.
If your caffeine is early, your drinks are early, your mornings are bright and your room is cold — and sleep is still harder than it should be — that’s where a formula is a reasonable next thing to try. Not a cure, and not a substitute for the four above.
One more piece of honesty about measurement, since most people now judge their nights by a wearable: your watch is good at telling sleep from wake and genuinely useful for trends and consistency. Its deep sleep number is an estimate produced by an algorithm from movement and heart rate, and in lab comparisons that estimate is mixed to poor. Judge a change by how you feel across a week, not by one night’s score.
STRAIGHT ANSWERS
Asked and answered.
What time should I have my last coffee before bed?
Caffeine's half-life averages about five hours, and a pooled analysis puts the cutoff for a standard cup at about 8.8 hours before bed. In a controlled trial, 400 mg taken a full six hours before bed cut objectively measured sleep by more than an hour while participants reported sleeping normally. Our two-week test: hold bedtime steady and cut caffeine at bedtime minus ten hours.
Does alcohol help or hurt sleep?
Both, in the worst order: it shortens the time to fall asleep at every dose tested, then fragments the second half of the night, delays REM, and drives the classic 2-to-4 AM wake-up as your liver finishes clearing it. Wearable data from over 4,000 adults found even one drink measurably reduced physiological recovery. Finish drinks three to four hours before bed.
Is my wearable's deep sleep number accurate?
Treat it as an estimate. Wearables are good at telling sleep from wake and genuinely useful for trends, but the deep-sleep figure is produced by an algorithm from movement and heart rate, and in lab comparisons that estimate is mixed to poor. Judge a change by how you feel across a week, not one night's score.
WHEN THE REPORT PUBLISHES
You’ve read the evidence. Decide when the report publishes.
Founding-list members receive the final label and the first batch’s lab report before the first purchase window. Free to join — no reservation, no charge.